TW201209753A - Method for image processing - Google Patents

Abstract

An image processing method includes the following steps. Firstly, input data including a number of original data are received. Then, converted emulation voltage signals are obtained by means of converting the original data. Next, at least a simulation circuit model, including at least a spatial data node, at least a diffusion node, and at least a connection device, is established. The at least a connection device is coupled a part of or both the at least a spatial data node and the at least a diffusion node. Next, a part or all of the converted emulation voltage signals are supplied to the at least a spatial data node to achieve voltage diffusion among the at least a spatial data node and the at least a diffusion node via the at least a connection device, so that at least a diffused emulation voltage signal is obtained on the at least a diffusion node. After that, processed image data are obtained according to the diffused emulation voltage signals.

Description

201209753 i wojyym f VI. Description of the Invention: [Technical Field] The present invention is a method for performing image processing via a circuit model simulation operation, and an embodiment includes an image related to depth data generation, image smoothing, and image resolution scaling. Processing operations. [Prior Art] In the current era of rapid development of science and technology, the stereoscopic multimedia _ system has gradually been valued by the industry. In general, in the application of stereoscopic video/video, how to compare two Dimensional (2D) images into two-dimensional (Three Dimensional '3D) stereo images and dual-view images (Stereo) Matching and other image processing technologies have always been the core technology of stereoscopic images that the industry is in urgent need of development. In the prior art, the 2D to 3D technology converts a traditional single-view image into a multi-view content, thereby providing a stereoscopic image content provided by the user; and the dual-view image comparison technology is first calculated based on the dual-view image. The depth and degree maps are used to generate multi-view images based on Depth Image Based Rendering 'DIBR. In general, the accuracy of depth data has a decisive influence on the quality of stereoscopic image data. Based on this, how to design an image processing method that can produce accurate and higher depth data is one of the directions that the industry is constantly striving for. SUMMARY OF THE INVENTION According to a first aspect of the present invention, an image processing method is provided, including

201209753 TW6399PA ^Steps. First, the input data including multiple original materials is received. Connect = to convert the raw material to produce multiple converted analog voltage signals. And then constructing at least one road model, including at least one spatial data node, =, a spatial data node, and at least a diffusion node, wherein t = smear then provides part or all of the converted analog voltage signal to:: a diffusion node, And diffusing = part or all to at least one diffusion node via at least one connecting element; = = obtaining at least one diffused analog voltage signal. After that, according to the second:, the political imitation voltage signal produces processed image data. According to the present invention, the second aspect proposes an image processing skirt, including receiving and converting m pseudo-units and control elements == round-up data of the starting data. The conversion unit pairs the original two-way model, the unit establishes at least - the analog electrical point, and the at least one connecting element is coupled to at least one of the spatial data sections and some or all of the two diffusion nodes. The control unit provides part or all of the conversion number to at least the "diffusion node", and all the two connections 70 pieces will convert part of the analog voltage signal or the at least - diffusion node, so that at least the diffusion node is obtained. Even pressure reduction. The ship unit turns at least - the pen spreads the simulated κ soul to produce processed image data. In order to better understand the above and other aspects of the present embodiment, the preferred embodiments are described below in detail with reference to the accompanying drawings, and the following detailed description is as follows: 201209753 [Embodiment] The image processing apparatus and method of this embodiment are The circuit model simulation operation is applied to perform related image processing operations. Please refer to FIG. 1 for a flowchart of an image processing method according to an embodiment of the present invention. The image processing method of this embodiment includes the following steps. First, as step (a), the input data Di including a plurality of original materials is received. Then, as in step (b), the original data is converted to generate a plurality of converted analog voltage signals V. And then, as in step (c), establishing at least one analog circuit module type, including at least one spatial data node, at least one diffusion node, and at least one connection component, wherein the at least one connection component is coupled to the at least one spatial data node and the At least one between the diffusion nodes. Then, according to step (d), part or all of the converted analog voltage signal v is supplied to at least one diffusion node, and through at least one connecting component: converting part or all of the analog voltage signal to at least one diffusion node, so as to At least one diffusion node obtains at least one diffused analog voltage signal v-diff. Then, as in step (e), the processed image data is generated according to at least one of the diffused imitation voltage signals • v—diff. Next, a series of embodiments will be taken to further explain the image processing method of this embodiment. First Embodiment The image processing method of this embodiment is applied in two dimensions (Tw0).

Dimensional ‘ 2D) Three Dimensional (3D) technology is used to generate depth distribution data based on initial depth data. Please refer to FIG. 2, which illustrates the input according to the first embodiment of the present invention.

201209753 TW6399PA information shows J. The actual image processing method generates depth distribution data DQ according to the input data μ. For example, input (4) Di (four) _ pen original data DiaDwu), ^ ^), ..., Di (mn), which is greater than 1 since _. The input data Di corresponds to the image data DV displayed in the display, and the image data DV includes the _ pen priming data muscle 1), 10, 2), ..., ^ (8) 'where the original data Diai)-Di(m, n ) corresponding to the pixel data K1,1)]^) of the face pixels displayed on the display. In this embodiment, the input data Di is the initial depth data corresponding to the image data Dv, and the values of the original data Di(l,l)-Di(m,n) of each of the pens respectively indicate the corresponding data of the subsequent pen data. Depth; depth distribution data Dq is the depth distribution data with higher fineness according to the initial/woodiness-bean material of the image data DV. Accordingly, the image processing method of the present embodiment is applied to generate three-dimensional (3D) stereoscopic image data according to two-dimensional (2D) image data and initial depth data, in other words, to perform 2D image content. Transfer processing of 3D stereoscopic images. ^ For example, the original data Di(l,l)-Di(m,n) respectively includes mxn pen 8-bit data, in other words, each original data Di(1, has a digit value between 0-255. When the original data Di(1, Djihn) corresponds to a larger digit value, it indicates that the corresponding halogen data 1(1,l)-I(m'n) has a shallow depth; when each original data Di(l) When 'l)-Di(m,n) corresponds to a smaller digit value, it indicates that the corresponding pixel data 1(1,1)-I(m,n) has a deeper depth. Please refer to Figure 3, A flow chart of the image processing method according to the first embodiment of the present invention is shown. First, as step (a), the initial depth data corresponding to the image material 201209753 1 material DV is received, and the mxn pen source depth data is used. As the mxn pen original data Di(l,1)-Di(m,n). Then, as step (b), the mxn pen original data Di(l,l)-Di(m,η) is converted to Corresponding to the mxn pen original data Di(l,1)-Di(m,n) respectively generate mxn pen-converted analog voltage signals SV(1,1), SV(1,2).....SV(m,η For example, the conversion step of step (b) is directly based on each piece of original data Di(l,1)-Di(m) [eta]) as the value of the digital voltage signal of each pen converted imitation SV (1, l) the digital voltage value -SV (m, η) of.

Then, as in step (c), an analog circuit model including at least one spatial data node, a J diffusion node, and at least one connected component is established. In an example of operation, the analog circuit model, as shown in Fig. 5, includes a sub-circuit model M (1, 1}, M(1, 2)..... 1^(111) having a similar circuit structure. , 11), which correspond to 111 < 11 original data 1) 1 (1, 1) - 1) 1 (111, 11), respectively. Since each mxn sub-circuit model has a similar circuit structure, next, only the sub-circuit model M(i' j) corresponding to the original data Di(i, D) in the analog circuit model is taken as an example to each of the analog circuit models M. The sub-circuit model uses (6)(10)^ as a step-by-step description, where ^^ and 卜 are respectively a natural number less than or equal to in and a natural number less than or equal to n. 钿 Refer to Figure 4, where 缯 '示依^ --- The circuit diagram of the road model M(i,j) is not invented in the first embodiment. The sub-circuit model M(i,j) includes the space nutrient node NS(i,j), the diffusion node_, 』·) , spatial data connection element _ this and two diffusion connection elements, the suppression ship, ..., version, the number, wherein the spatial data connection element Rs and the diffusion connection element age service, for example, resistance_component 4 (four) (6) towel, empty four face In the spatial data node NS (i, D and expansion 201209753

Between i wwyypA, one of the z diffusion connection elements RD1_RDz is coupled to the diffusion node ND(i,j), and the other end is coupled to a diffusion node of another sub-circuit model in the analog circuit model M. In the example of Fig. 4, ζ is equal to 4; and in step (c), the other ends of the diffusion connecting elements RD1-RD4 are respectively coupled to the sub-circuit mode = Μ(ι 1' j), M(i, Diffusion nodes ND(i-l' j), ND(i, jl), ND(i, j + 1) and ND in ji), M(i, j + l) and M(i + 1, j) (i + l, j). Similarly, in step (c), all of the mXn diffusion nodes ND(11)_ND(mn) in the mxn sub-circuit models M(M)-M(m,n) are connected to each other via corresponding diffusion interfaces. The connection is such that each sub-circuit model Μ( 1 ' 1) M(m, η) in the analog circuit model 串联 is connected in series to form a node and a resistor network, as shown. For example, the spatial data diffusion connection elements in all analog circuit models, i.e., the resistance values obtained by (1) are substantially the same as the values given by the user. For example, in the analog circuit model M (i, D, the resistance values %^ of the z diffusion connection elements RD1 - RDz satisfy: ^, "々 is a predetermined parameter; Ct is the original data Di(i,]·) corresponding The color information of the 昼素=贝=; Cn is the expansion of the connection of each diffusion connection element: lang (ie, '1-1, j), ND (i, ji), ND (i, j + 1) and _ On D(i + l, j)), the color of the pixel data corresponding to each piece of original data:: For example, the color information of the element data. The enthalpy can be obtained from the sum of the absolute values of the sub-halogen data of the respective colors in the corresponding bain. 201209753 1 wwjyyr / v , then as step (d), the corresponding imitation voltage signal SV(1, l)-SV(m,n) corresponding to the mxn pen original data Di(l,1)-Di(m,n) Provided to mxn spatial data nodes NS(1,l)-NS(m, η), respectively, thereby driving mxn sub-circuit models via voltage diffusion operations between spatial data connection elements and diffusion connection elements in the analog circuit model M (l, DKm, η) voltage level redistribution occurs, so that the diffusion node NDd, respectively, obtains the mxn pen diffusion analog voltage signal SVD (1, 1), SVD (1, 2) ..... SVD (m , η).

Then, as in step (e), the depth profile data D〇 is generated based on the mxn pen spread analog voltage signal SVD(1,1)-SVD(m,n). In the image processing method of the embodiment, only the step (a) directly uses the digital value of each of the original data Di(l, 1)-Di(m, n) as the converted analog voltage signal SV (1). The case of the digital voltage value of l)-SV(m, n) is taken as an example. However, the image processing method of the present embodiment is not limited thereto. In other examples, when the image data is the dynamic video data, the image processing method of the embodiment may also perform the conversion of the analog voltage signal X according to the original material D1 (x, y) according to the next financial scream. Operation: SV(x, y) = yx Dipre (x, y) + (l. r)x Di(Xj y) where X and y are less than or equal to, respectively, the natural number and less than or equal to = natural number 4 It is pre-compared with the parameter; Dipre(xy) is corresponding to the previous frame time (Frame Time); j is less than the number of priests and his depth data;; medium = image corresponds to the nature of 昼素二number. Through the foregoing operations, the shadow processing method of this embodiment may be called strong depth distribution data D. The depth contrast. 201209753

TW6399PA In another example, the image processing method of the present embodiment can also achieve the effect of emphasizing the depth of the moving object by superimposing and strengthening the voltage on the spatial data node Ns (which originally has the converted analog waste signal SV). . For example, the image processing method of the present embodiment performs the operation of superimposing the enhanced voltage on the spatial data node according to the following equation: SV(x, y) = Di(x, y) + min(^ ψρΓε (x , y) _ C<:ur (x> y)|) where "^(心和^^^-'匕- is the superimposed voltage value to be superimposed on the converted imitation voltage signal... is a predetermined parameter; 4 The upper limit value of the enhanced voltage value; CPre(x, y) and Ccur(x, y) are the color of the pixel data at the (x, y) position of the previous frame time and the current frame time, respectively. In the embodiment, although only the sub-circuit model M(l, l)-M(m, n) having the same number of Di(m, n) as the original data Di(l'l) is included in the analog circuit model M, And respectively receiving the original data Di(l,l)-Di(rn,n)f^^#^^tMm with the data nodes NSaU-NSOn'n) on the sub-circuit model M(M)-M(mn) The case of SV(1, 1)-SV(mn) is taken as an example. However, the image processing method of this embodiment is not limited thereto. In other examples, the analog circuit model generated by the image processing method in the step (c) may also include a sub-circuit model whose number is not equal to the number of original data; correspondingly, the user may also make the data node of the part to be connected. (Floating) or discarding some of the original data means to input some or all of the original data DiCl'D-DiCm, !!) into the analog circuit model ' 'by taking similar operations to obtain the processed image data. Second Embodiment The image processing method of the present embodiment is applied to the dual-view image comparison 201209753 1 wvjwrn (Stereo Matching) technique for generating depth distribution data based on the first and second viewing angle image data. Referring to Figure 6, there is shown a schematic diagram of input data not according to the second embodiment of the present invention. In this embodiment, the input data Di is a parallax data corresponding to the first-view image negative material DvL and the second viewing angle image data DvR, and the depth distribution data Do′ is corresponding to the first or second viewing angle image data DvL and DvR. Depth distribution data. Accordingly, the image processing method of the present embodiment is applied to generate corresponding depth distribution data based on the parallax data of the first and second viewing angle image data DvL and DvR Φ, in other words, to perform dual-view image comparison operation. In more detail, the image processing method of the second embodiment is different from the image processing method of the first embodiment in that the step (a,) further includes sub-steps as shown in Figs. 7A and 7B. First, as step (ai), the first view image data and the > one view image data DvL and DvR are received. For example, the first and second viewing angle image data DvL and DvR are image data corresponding to the left eye angle and the right eye angle, respectively. • Then, as in step (a2), determine the w horizontal disparity values Dxl, Dx2....

Dxw, and when the first view image data DvL has the kth horizontal disparity value Dxk with respect to the second view image data DvR, it is found that the first view image data DvL is different from the first view image data DvR. Disparity DIS_k, k is the index of the image comparison window, and its value is a natural number greater than or equal to 1 and less than or equal to w. For example, the first original dissimilarity data DIS_k includes mxn pen original prime dissimilarity data DIS(1,1, Dxk), DIS(1,2, Dxk).....DIS(m,n, Dxk), where step (a2) is operated by the following equation to find the first original dissimilarity 201209753

TW6399PA Negative material DIS_k each mxn pen original 相 相 dissimilarity data DISCI, l, Dxk) - DIS (m, n, Dxk): '

DvR- (x, y) = I [DvR(x, y) + DvR(x -1, y)]

DvR+ (x,y) = I [DvR(x,y) + DvR(x +1,y)]

DvRMi„ (χ» y) = Min(DvR' (x, y), DvR+ (x, y), DvR(x, y))

DvRMax (x. Υ) = Max(DvR- (x, y), DvR+ (x, y), DvR(x, y)) DIS(x,y,Dxk) = Max(0,DvL(x,y) -DvRMM(x.Dxk,y), DvRMin(x.Dxky)_DvL(x5^ where X and y are respectively a natural number less than or equal to m and a natural number less than or equal to n. Then, as in step (a3), The first original dissimilarity data DIS_k is used as the input data Di' 'where mxn pen original data Di, (11) Di, (mn) is the first-original dissimilarity data DIS_k, respectively corresponding to the individual elements Ul'l)-I(m,n) raxn pen first original elemental dissimilarity data DIS(1,1,Dxk)-DIS(m,n,Dxk). The image processing method of the present embodiment is After the step (a3), the steps corresponding to the steps (b)-(d) of the third figure are performed correspondingly: (匕,), the original data of each m xn pen is converted to generate the claw Xn pen-converted pseudo-voltage respectively. Signal not, (1,1)-SV, (m,n); (c,) corresponding to m Xn pen original data Did, D-DUm'n) generate analog circuit model Μ, including diffusion node ND, ( 1,1)-Europe (ιη, η) and corresponding diffusion connecting elements are connected to form a node and a resistor network; and (d,) will convert the analog The signal (1,1)-8¥'(111,11) is provided to 111><11 spatial data nodes S (1, 1)-NS (m, η), and mxn sub-circuit models ^1 (1'1) 扩散'(111,11) diffusion node 仙'(1,1)_ rib, (111,11) respectively obtain mXn pen diffusion analog voltage signal SVD, (1,1,Dxk)-SVD, (m, n, Dxk). 12 201209753

1 I For example, 'the sub-circuit model M of this embodiment, (l, l)-M, (mn) is different from the sub-circuit model milk (1) bin (10) in the first embodiment) in the sub-circuit model M, (i , j) The resistance value %_ of a diffusion connecting element RDl'-RDz' satisfies: _ a

^diffuse = ----T where or, h is the predetermined parameter; Gs(x, y) is the gradient value of the first-view shirt like the material DvL after the smoothing operation, and the pixel corresponding to each of the original materials # 5 The material color of the material, Cn is the color information of the pixel data corresponding to the original data received by the sub-circuit model connected by the diffusion connecting element RD1, -RDz'. After the step (d) and before the step (e), the image processing method of the embodiment further includes steps (f), (g) and (h). In step (f), it is determined whether the value k is rounded up between 1 and the natural number between the dissimilarity search window parameters w; if not, it indicates that the image processing method of the embodiment has not been applied to all w levels. The disparity values Dxl, Dx2, ..., Dxw find out the corresponding first original pixel dissimilarity data DIS_l-DIS_w (the corresponding transposition reference voltage signal and the diffusion imitation voltage signal). According to this, step (g) is performed to set the parameter k to a natural number between 1 and the dissimilarity finding window parameter w that has not been rounded, and repeat the operations of steps (al)-(a3) to Find a first original dissimilarity data DlS_k corresponding to the next horizontal disparity value. After finding the next first original dissimilarity data, the image processing method of the present embodiment also repeats the operations of steps (b,)-(d') to generate corresponding to the next first original dissimilarity. The analog circuit model M' of the data, and the mxi1 pen diffusion imitation 13 corresponding to the next first original dissimilarity data is obtained 2012 201253

TW6399PA pressure signal SVD' (1, l, Dxk) _SVD, (m n Dxk). The above is for all the w horizontal disparity values Dxl, Dx2.....Dxw to find the corresponding part of the w original first prime dissimilarity data DIS_J-DIS_jv = or all the processing flow can also be parallel The processing method is carried out. When the value k between the 1 and the dissimilarity looking window parameter w is completed, it indicates that the image processing method of the embodiment has been found for all W 7& Corresponding to w, the original dissimilarity data DISJ-DIS-w, in other words, correspondingly find out the first original book of wxmxn pen, s, i, dx1h> hDx1)

DIS(1, 1, Dx2)-DIS(m, n, Dx2).-DIS(l, 1, Dxw)-DIS(m, n, D XW). According to the image processing method of this embodiment, step (h) is performed to find the w-split imitation voltage sfl number corresponding to each of the original data Diai)-Di(m n). In the mxn pen original data j)in η.Λ, 山, 笔原., () (m, n) in the (i, j) imitation electric Υ « I1 as an example, its system corresponds to the following W pen Diffusion side, pressure signal SVD, aj, Dxl), SVD, (ij, Dx2), (i, J, Dx3), ..., SVD, (i, j, Dxw). Correspondingly, the step of the image processing method of the embodiment (including the 簦 to find each 原始 笔 原始 原始 原始 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 It It It It It It It It It It It It It 1'n_SVDmin〇n,n), to find out the depth score ==, respectively, with the original data Di, (u) money miscellaneous ^ output pixel dissimilarity data D 〇, (U), D 〇, a2), ..., Do, (mn). The original data Diai) The first (6) of the Di (m, n) pen original capital Dl (1'J) For example, the operation of the step (8) can be the following equation table 201209753 i Wo^vy^A . Show:

Do(i, j) = SVDmin (i, j) = ^ rmnSVD-(i5 j, Dxk), k = 1,2,..., w According to this, the original data Diaj) can be obtained correspondingly. , (i,]·). In another example, step (e) further includes step (e2) to diffuse a quadratic function curve according to the minimum diffusion front and back of the lowest diffusion analog voltage signal, and thereby quadratic The most j of the function curve is used to find the output pixel dissimilarity data Do (l, l)-Do'(m, n) with accuracy below the decimal point. For an example of operation, the first data Di(i, j) corresponds to the output 昼 相 dissimilarity data d. , ('丨 is the lowest diffusion imitation voltage signal SVDmin(i,]·)) for the corresponding horizontal view

The diffusion of Dx5 is like a voltage signal, U, j, 5). (4) (4) The diffusion-like voltage signals SVD (1, j, 4), SVD' (i, 5), and SVn, f corresponding to the horizontal disparity values Dx4, Dx5, and D, J Jux. λ, 1 (8) and MD (1, J, 6) are substituted into the following equations to find the parameters a, b and 0 correspondingly: ax2 + bx + c = y which corresponds to the original data D i Π Π > M , a * Bayer Dl (1'J) output 昼 相 dissimilarity data

Do (1, j) is equal to: D 〇 (i, j) = according to this, according to the operation of the foregoing steps (a,) _ (e,), (f) _ (h), the lining processing method of this embodiment Correspondingly according to the first-view image data DvL (for example, the left-eye view image) relative to the second-view image data

The first original dissimilarity data of DvR (for example, the right eye viewing material data) DISJ-DISj is used to generate the depth distribution data D〇. 201209753, TW6399PA ' In the present embodiment, the depth is generated only by the first view angle image data DvL of the image processing method of the present embodiment with respect to the first original dissimilarity data DIS_l-DIS_w of the second view image data DvR. The case of distributing the data D〇 is taken as an example. However, the image processing method 1 of the present embodiment is not limited thereto. In another example, the image processing method of the embodiment is further V, according to the operation similar to the foregoing embodiment, correspondingly according to the second view image data DvR (for example, the right eye view image data) relative to the first view image data. The second original dissimilarity data DIS'-l-DIS'_w of the DvL (for example, the left-eye viewing angle image data) is used to generate the depth distribution data Do. In still another example, the image processing method of the present embodiment can compare the depth distribution data Do' and the second perspective image data DvR of the second perspective image data DvR with respect to the first perspective image data DvL. The consistency of the depth distribution data Do" of the DvL image data is used to improve the accuracy of the depth distribution data. For example, for the first view depth distribution data Do' and the second view depth distribution data D〇, the output 昼 相 dissimilarity data in which the following equation conditions are satisfied is retained: lD〇' (X, y) - Do" (χ - Do- (X, y), y)| s!. For s, it is only the depth distribution data D〇, which is consistent with the output pixel dissimilarity data. _ For no The output pixel disparity data that satisfies the above equation can be filled with any material like I叩ainting technique. This example uses its corresponding round-off morphological data to its left and right sides. The position is the closest and satisfies the minimum value of the output pixel dissimilarity data of the foregoing equation (that is, is retained). In this embodiment, although only the image processing method is used in step (b,), 201209753 The sub-circuit model ^(3)-the heart (8) has the circuit structure as shown in the (10) as an example. The image processing of the present embodiment is not limited thereto. In another example, in the image processing method Established mxn sub-circuit models M"(11) . . . . ^ The data node and the time data diffusion connection Yuan Yuan m'n ° ru, called, n) in the first (i analog electric cap ' 模拟 an analog circuit model Mn (i, j), which is shown in Figure 3 The circuit structure is different in that it further includes a time data node NT(i, j) and a time data diffusion connection component, as shown in Fig. 8. The time data diffusion connection element RT (i, D is light Between the time data node NTU'j) and the diffusion node mail (1), and the resistance value of the time data diffusion connecting component RT(i,j) is called ime: where Α, σ are predetermined parameters, and L is the original data at present The color information of the corresponding pixel data in the frame time (for example, the number of each of the pens in the first-view image data), 'C<« is the corresponding data of the original data in the previous frame time. According to the image processing method of the present embodiment, the depth distribution data D〇' may be generated by referring to the color information corresponding to the front and back of the different frames. In addition, 'such as monocular cues information may also be applied ( M〇n〇cular Cues) (for example, linear perspective (Li Near perSpective) to assist in the operation of the depth distribution data Do'. In addition, semi-gi〇bai and global alignment mechanisms, such as dynamic programming or trust diffusion mechanism (Bei ief pr〇) Pagati〇n) can also be 17 » 9 201209753

The TW6399PA is applied to the image processing method of this embodiment to improve the alignment accuracy of the first and second viewing angle image data DvL and DvR. In this embodiment, the image processing method directly uses the first and first view image negative materials DvL and DvR to perform the correlation comparison operation as an example. However, the image processing method in this embodiment does not Limited to this. In other examples, in order to simplify the overall data calculation amount, the image processing method in this embodiment may perform a resolution reduction operation on the first and second perspective image data DvL A DvR before the operation is performed; in other words, 'According to the first and second viewing angle image data DvL A DvR纟 after the resolution reduction, the comparison operation is performed, and the depth distribution data with lower resolution is obtained. After that, the depth distribution data is enlarged to obtain the depth of the same resolution when the amount of data calculation is greatly reduced. For example, the foregoing operation for amplifying the depth distribution data in the embodiment may be implemented through the flow of the image processing method as shown in FIG. 1 : there is a depth distribution data to be amplified, for example, Enlarged to a resolution equal to s, xt, = large ice scores, where s, t, s, and t are large W and S and t are respectively full: S < S, and t < t,. In this operation example, the image processing method of this embodiment is generated by ', = node and s, xt, and diffusion nodes. Including sxt, the scatter operation operates on s, xt, and the diffusion node gets the model' and spreads the voltage signal via voltage. According to this, the pen-diffusion node diffusion image processing square wire can perform the data analysis and disambiguation operation correspondingly via the foregoing electromotive; 201209753 1 w®^yyr/\ . Third embodiment

The scene of the present embodiment> image processing method generates depth distribution data corresponding to the image based on the data input by the user. The difference between the first embodiment and the second embodiment is that the image processing method of the embodiment further provides a user interface to receive a user operation event provided by the user; and the image processing method of the embodiment is provided by the user. The user operates the event to selectively increase or decrease a part of the nodes and the diffusion connection elements of the mxri sub-circuit model, or selectively set the resistance values of the voltage signals and the diffusion connection elements corresponding to the spatial data nodes in the sub-circuit model. In addition, the image processing method of the present embodiment respectively drives the voltage level re-distribution corresponding to the analog circuit model to respectively obtain the user-controlled diffusion-like voltage signal on the diffusion node of the analog circuit model. For example, the user interface provides a segmentation image processing tool and a brush tool. The image segmentation ^ is configured to selectively perform an object segmentation operation on the input data in response to a user-triggered user operation event to find out the object information in the input data; the brush tool allows the user to selectively assign Numerical information is given to the corresponding input data. The image processing method of the present embodiment can further increase or decrease in the operation step (e) of the pseudo circuit model by referring to the aforementioned # 点 或 point or the diffusion node, or set the resistance value of the diffusion connection element and the vacant connection element. a s =Bein In an operation example, in the operation of establishing an analog circuit model (c), the image processing method first refers to the object allocation information to the gentleman = the sudden component to set the resistance value. When the object distribution f refers to the same object segmentation, the image processing method of this embodiment ^ = 201209753

TW6399PA * tl> The same method as in the first embodiment is used to set the resistance value of the diffusion connecting element during the period; when the object allocation information indicates that the two diffusion nodes belong to different object segments, the image processing method of the embodiment Correspondingly, the resistance value of the diffusion connecting element between them is set to a maximum value to ensure that the diffusion between the different object segments has a lower voltage diffusion condition. Then, in the operation step (C) of establishing the analog circuit model, the image processing method then establishes the data node and the data connection component according to the numerical data assigned by the user (corresponding to the specific diffusion node); For a diffusion node that does not assign any original data, the image processing method does not perform related data node and data connection component establishment operations. Accordingly, the image processing method may refer to the reference object segmentation information and the user's numerical data assignment information in step (c) to establish a corresponding analog circuit model' to perform a corresponding image processing operation. Fourth Embodiment The image processing method of the present embodiment is applied to an image smoothing application to generate smoothed image data for input image data. The image processing method of this embodiment is configured to generate image smoothing data Do" according to the input data D i ”. For example, the wheeled data Di" includes mxn pen sputum data 1 corresponding to mxn pixels (1, 丨^^", which includes the first pixel data lsubl(l,l)-lsubl(mn), and the image smoothing data Do" includes the mxn pen smoothing sub-sputum data Isml corresponding to this mxn element l, 1) - Isml (m, η). 201209753 » The month refers to Fig. 9, which is a flowchart showing an image processing method according to a fourth embodiment of the present invention. First, as in step (a), the first data of the mxn pen, Isubl(l,l)-Isubl(m,n), is received and used as input data. Then, as in step (b), the first morpheme data 13111)1(1,1)-13111:)1(111,11) of each mxn pen is converted to respectively generate an η-like converted analog voltage signal. Sv(l,l)-SV(m,n). Then, the analog circuit model M' " is generated as the step (c) ' corresponds to the mxn pen first time data Isubl(l,l)-isubl(m,n). For example, say 'analog circuit model M,' includes mxn sub-circuit models M'"(l'l)-M'"(n,m), each mxn analog circuit model river'"(1,1) -1^'"(11,111) includes a data node milk, a diffusion node rib, a data diffusion connection element RS, and X diffusion connection elements rD1_RDx, and the data diffusion connection element RS is coupled between the data node NS and the diffusion node ND. One end of each of the X diffusion connection elements RD is coupled to the diffusion node nd, and the other end is connected to another sub-circuit model of mxn sub-circuit models M, n(l, 1)-M, n(m, n), wherein X is a natural number. Lu then, as in step (d), converts the first imitation data isubl(l,l)-Isubl(m,n) corresponding to each mxn pen into a pseudo-voltage signal 3¥(1,1)-8¥(111 , 11) provide data node milk to drive 111) <11 sub-circuit models M' "(1,1)-M' "(m,η) voltage level reallocation, to mxn The circuit model M, "(l,1)-M,"(n,m) diffusion node ND respectively obtains mxn pen diffusion-like voltage signal SVD(1, after step (e)' according to mxn pen diffusion simulation The voltage signal SVD(l,l)-SVD(m,n) produces image smoothing data including the mxn pen smoothing sub-pixel data Isml(l, 1)-Isml(m, η). 201209753 TW6399PA * ' ' In one case, each mXri pen food data 1 (1,1)-1 (m,η) contains, for example, the second halogen data and the third halogen data separately.

Isub2(l,1)-Isub2(m,n) and Isub3(l,1)-Isub3(m,n), and the image processing method of the present embodiment is further, for example, via the aforementioned steps (a)-(e) The similar operation steps are performed to find the corresponding smoothed secondary data Ism2(l,1)-Ism2(m,η) and 1fine 3(1,l)-Ism(m,η). Fifth embodiment

For example, the foregoing operation of amplifying the depth distribution data in the present embodiment can be implemented via the image processing method flow steps as shown in FIG. 1. In an operation example, the depth distribution data to be amplified is, for example, sxt pen original data, and is to be enlarged to a resolution equal to s, xt, and the amplified depth distribution data 'where s, t, s' and t are A natural number greater than 1, and s and t satisfy: s<s' and t<t', respectively.

In this example of operation, the image processing method of the present embodiment obtains s by generating an analog circuit model including sxt data nodes and s'Xt' diffusion nodes, and operating at S' Xt' diffusion nodes via voltage diffusion. Xt,' pen spreads the point spreads the imitation voltage signal. Accordingly, the image processing method of the present embodiment can also perform the resource amplification operation correspondingly via the aforementioned circuit β. ^ Resolution According to this, by selectively setting the number of expanded data nodes in the analog circuit model, the image is calculated by the real image and the resolution scaling operation of the image data. In the image processing method described in the embodiments of the present invention, 22 201209753 can be realized and recorded in a corresponding computer readable medium. In this manner, the user can use the computer processor to access the computer readable memory to perform the image processing method according to the foregoing embodiments of the present invention according to the computer readable program stored in the towel. For example, the image processing method of the foregoing embodiments of the present invention can be implemented by the image processing apparatus 1 shown in the figure. In more detail, the image processing apparatus 1 includes an input port; a control unit 1 and a control unit it 40. In the element 2, the analog unit 40 is a software module, the '2' is replaced by the unit, and the control unit executes the phase (4). The code is changed by the processor, and the input and reception include multiple The input information of the original data Di. = 2: The original data is converted to generate a plurality of conversion simulations including at least L*nt^3G to establish a corresponding analog circuit model, wherein the package to / the resource node, at least one of the Mingke βs, ± Pieces. The control unit 40 converts the imitation "(4) to >, a connection element to at least the diffusion node, and provides a ψ /V of part or all of the number: the connection element will convert some or all of the imitation voltage signal. Diffusion to at least one diffusion node to obtain at least - pen = &5; 5 ^ unit according to at least - pen imitation ^ scattered voltage signal V - sen. Control image data 0., voltage ring v-dlff generated processing shadow

The Mj month description of the embodiment is related to the image processing method. Compared with the method, the image processing product of the embodiment of the present invention has the advantage of the stereo image (4); in comparison with the conventional image smoothing method, the image of the embodiment of the present invention is compared with the image of the embodiment of the present invention. The processing method has the advantage of effectively smoothing the input image. In summary, although the patent specification has been disclosed in the above preferred embodiments, it is not intended to limit the disclosure. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the disclosure. Therefore, the scope of protection of this disclosure is subject to the definition of the scope of the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a flow chart showing an image processing method according to an embodiment of the present invention. Fig. 2 is a schematic view showing input data in accordance with a first embodiment of the present invention. Fig. 3 is a flow chart showing an image processing method according to the first embodiment of the present invention. Fig. 4 is a circuit diagram showing a sub-circuit model in accordance with a first embodiment of the present invention. Fig. 5 is a circuit diagram showing an analog circuit model in accordance with a first embodiment of the present invention. Figure 6 is a diagram showing the input data in accordance with the second embodiment of the present invention. 7A and 7B are flow charts showing an image processing method according to a second embodiment of the present invention. Figure 8 is a circuit diagram showing a sub-circuit model in accordance with a second embodiment of the present invention. 24 201209753 FIG. 9 is a flow chart showing an image processing apparatus according to a fourth embodiment of the present invention. Figure 10 is a block diagram of an image processing device in accordance with an embodiment of the present invention. [Description of main component symbols] DV: Image data Di: Input data • Ι(1'1)-Ι(πι,η), Il(l,l)~l1(m,n), I2(l,l)- I2(m,n): Alizarin data Di(l,1)-Di(m,n): original data M(i, j), M" (i,j): analog circuit model NS(1'l) -NS(m,n): spatial data node ND(1,l)-ND(m,η): diffusion node RS: spatial data diffusion connection element RDl-RDx: diffusion connection element φ Dxl-Dxw: horizontal disparity value NT: time data node RT: time data diffusion connection component

DvL, DvK: first-, second-view image data 1: image processing device 10: input unit 20: conversion unit 30: analog unit 40: control unit 25

Claims (1)

201209753 TW6399PA VII. Application for Patent Park: 1. An image processing method comprising: receiving a plurality of original data; converting the original data to generate a plurality of converted analog voltage signals; establishing an analog circuit model, the analog circuit The model includes at least a material node, at least a diffusion node, and at least a connection element, wherein the to the first small connection element is coupled to some or all of the at least one data node point; Providing some or all of the converted analog voltage signals to the at least one data node, and converting the plurality of converted analog voltage signals to the at least one connecting element: part or all of the voltage signals are diffused to the at least one-diffusion node, so as to At least the diffusion node obtains at least one diffusion-like voltage signal '; and generates at least one processed image data according to the at least one diffusion-like voltage signal. 2. The image processing method according to claim 1, wherein the original data is depth information, and the depth information is related to an image data. 3. The image processing method of claim 1, wherein the original material is depth information related to the received user input data. 4. The image processing method according to claim 1, wherein the connecting element is an element having a resistance characteristic. 5. The image processing method of claim 4, wherein the resistance value of the connection element is related to a color information in an image data. 6. The image processing method of claim 4, wherein the resistance value of the connecting component is related to a user input data. 7. The image processing method according to claim 4, wherein the resistance value of the connecting element is corrected by a smoothed image gradient. 8. The image processing method according to claim 1, The data node is related to the image pixel distribution or related to the frame time. 9. The image processing method of claim 1, wherein the original material is color information of an image. 10. The image processing method of claim 1, wherein the step of receiving the plurality of original data further comprises: receiving an initial depth data as an input data, wherein the initial depth data includes a plurality of pixels corresponding to the plurality of pixels. The original information. 11. The image processing method according to claim 10, 27 201209753 TW6399PA * - wherein the step of establishing an analog circuit model further comprises: providing a plurality of the at least one diffusion node in the analog circuit model, corresponding to each The at least one connecting node includes X diffusion connecting elements, one end of each of the X diffusion connecting elements is coupled to the corresponding at least one diffusion node, and the other end is coupled to the at least one diffusion node Another diffusion node in which X is a natural number. 12. The image processing method of claim 1, wherein the step of receiving the plurality of original data further comprises: (al) receiving a first view image data and a second view image data; (a2) finding out When the first view image data has a k-th horizontal disparity value Dxk relative to the second view image data, the first view image data and the first view disparity data of the second view image data (Disparity) And k is a natural number greater than or equal to 1 and less than or equal to w; and (a3) using the first original dissimilarity data as the input data, wherein the original data is the first original dissimilarity data The first original elemental dissimilarity data corresponding to the plurality of elements of the plurality of elements. 13. The image processing method of claim 12, wherein the step of establishing an analog circuit model further comprises: providing a plurality of the at least one diffusion node in the analog circuit model, corresponding to each of the at least one diffusion node The at least one connecting component includes X diffusing connecting components, one of the X diffusing connecting components is coupled to a terminal of 28 201209753 IW\iDyyrt\ , and the other end of the diffusion defect is coupled to the diffusing node, where X is a natural number. Where = 2 applies for the image processing method described in item 12 of the Patent Park, between: =::rr dissimilarity looking for the window parameter - the (8)-(10)" η column or the parallel processing method repeats the step as the input Data; Gongli first-primary phase μ degree data, wherein the at least-------the original dissimilarity is generated for generating data corresponding to the analog circuit model and obtaining corresponding/the first original dissimilarity data Repeat execution. g electric ridge signal, the relevant steps are correspondingly included, including the image processing method described in the stomach patent _ 14 And corresponding to the natural number between the other at least one of the at least one diffusion node, the dissimilarity finding window parameter, and the corresponding one to the at least one of the diverging nodes.扩散 夕 一 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散 扩散The money signal finds the processed image data. The method of image processing according to claim 15, wherein the step (el) further comprises: determining whether the minimum diffusion-like voltage signal corresponds to a reference diffusion-like voltage signal, Verifying the minimum spread analog voltage signal, wherein the reference diffused analog voltage signal is related to the dissimilarity data of the second view image data relative to the first view image data; and when the lowest spread analog voltage signal corresponds to When the reference diffuses the analog voltage signal, the minimum diffused analog voltage signal derived by the reference diffusion analog voltage signal is used as the processed image data. The image processing method of claim 12, wherein the step of establishing an analog circuit model further comprises: providing a time data node in the at least one data node, corresponding to the time data node, the at least one The connection component includes a time data connection component coupled between the time data node and the at least one diffusion node. 18. The image processing method of claim 1, wherein the step of establishing an analog circuit model further comprises: selectively determining the at least one data node in the analog circuit model in response to a first user operation event And the number of the at least one diffusion node and the at least one connection component; in response to a second user operation event, selectively setting a voltage signal corresponding to the at least one data node in the analog circuit model and the at least one connection component resistance. The method of image processing according to claim 1, wherein the step of receiving the plurality of raw materials further comprises: receiving a plurality of pixel data corresponding to the plurality of pixels, and using the plurality of pixels The sub-small data is used as the original data corresponding to the pixels; wherein the processed image data generated by the original data of the pseudo-pixel data is image smoothing processing video material. The image processing method of claim 1, wherein the step of receiving the plurality of original materials further comprises: receiving a first frame data, and using the first frame data as the input data, The first frame data corresponds to a first level of resolution; wherein the processed image data generated by the input data that is substantially the first frame data is one of the resolutions corresponding to a second side Image data after image scaling processing. The image processing method of claim 1, wherein the step of establishing an analog circuit model further comprises: providing a spatial data node in the at least one data node, corresponding to the spatial data node, the at least A connection component includes a spatial data connection component coupled between the spatial data node and the at least one diffusion node. 22. An image processing apparatus comprising: 31 201209753 TW6399PA *, - an input unit for receiving an input data, the input data comprising a plurality of original data; and a conversion unit for converting the original data, Generating a plurality of converted analog voltage signals; an analog unit for establishing an analog circuit model, the analog circuit model including at least one data node, at least one diffusion node, and at least one connecting component, wherein the at least one connecting component is coupled to the And at least one data node and part or all of the at least one diffusion node; and a control unit, providing a part or all of the converted analog voltage signals to the at least one diffusion node, and via the at least one connection component And diffusing part or all of the converted analog voltage signals to the at least one diffusion node, so that at least one diffusion-like voltage signal is obtained by the at least one diffusion node, and the control unit generates a processing according to the at least one diffusion-like voltage signal. video material. 23. The image processing device of claim 22, wherein the input unit receives an initial depth data and uses the initial depth reference material as the input data, wherein the initial depth data includes a plurality of corresponding depth data. The original data of Russell. The image processing device of claim 23, wherein the simulation unit provides a plurality of the at least one diffusion node in the analog circuit model; wherein, corresponding to each of the at least one diffusion node, the at least one The connecting element comprises X diffusion connecting elements, 32 201209753 - end handles of the respective X diffusion connecting elements are connected to the corresponding at least - diffusion nodes, and wherein X is from the other of the at least - diffusion nodes _ Sanjie 35 is the number. The image processing device of claim 22, wherein the receiver is configured to receive a first view image data and a first view image data; and an operator to find the first view image relative to the image When the angle image data has a k-th horizontal disparity value 该, the first image data and the second (four) image (4)--the original dissimilarity ^ = (DiSpanty), k is greater than or equal to 丨 to, or Equivalent to w = Ran, ', the data processor uses the first original dissimilarity data as the input corresponding to the plurality of pixels of the plurality of pixels - the original pixel dissimilarity 26. #申请专利范围第25 The image processing device of the present invention, wherein the canister unit provides a complex diffusion node in the analog circuit model; v, corresponding to each of the at least one diffusion node, the at least one connected-X diffusion connection elements, each of the X The diffusion connecting element should be the at least one diffusion node, and the other end is reduced to: > to > another diffusion node in the diffusion node, where X is a natural number. The image processing device of claim 25, wherein the control unit sets the value k to a natural number between 1 and the dissimilarity finding window parameter w, and correspondingly Driving the input unit to find and input the first original dissimilarity data as the input data in a sequence or parallel processing manner, wherein the conversion unit and the control unit perform corresponding operations to generate corresponding to the next And the analog circuit model of the first original dissimilarity data and the at least one diffused imitation voltage signal corresponding to the next first original dissimilarity data. 28. The image processing device of claim 27, wherein when the value k corresponding to the natural number between the dissimilarity finding window parameter w is completed, the control unit corresponds to The at least one diffusion node finds the at least one diffused analog voltage signal, and finds the processed image data according to a minimum spread analog voltage signal having a minimum voltage value among the at least one diffused analog voltage signal. The image processing device of claim 28, wherein the control unit further determines whether the minimum diffusion-like voltage signal corresponds to a reference diffusion-like voltage signal to verify the minimum diffusion-like voltage signal, The reference diffusion-like voltage signal is related to the dissimilarity data of the second view image data relative to the first view image data; wherein, when the minimum diffusion-like voltage signal corresponds to the reference diffusion-like voltage signal, the control The unit derives the minimum diffusion-like voltage signal of 2012 201253 53 A »» ΓΛ. t as the processed image data by using the reference diffusion-like voltage signal. The image processing device of claim 25, wherein the simulation unit provides a time data node in the at least - data node, corresponding to the time data node, the at least one time connection, the data connection component, The time data connection component is lightly coupled between the data node and the at least one diffusion node. Secondly, the image processing device according to Item 22, the soldier T-child control is responded to by a first mosquito, the mold (four) road model money, the at least the "fest node selection = the number of connected components" Response: "; less: the voltage signal corresponding to the spatial data node and the resistance value of the at least one piece.夕连接7C 0Z. (4) The image processing of the dream information according to item 22 of the patent scope, and the data of the second number to the plural number of digits, and the corresponding data to the elements These, which are based on the actual Zeng Bu Li # ^ > 产生 产生 产生 产生 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该data. ', the image processing device after the smoothing of the shirt image| 33. The image processing device according to claim 22 of the patent application scope, 35 201209753 TW6399PA 1 1 wherein the input unit receives a first frame data, and The first frame data is used as the input data, and the first frame data corresponds to a first facial image resolution; wherein the processed image data is generated according to the input data that is substantially the first frame data. The image data after scaling is processed to correspond to one of the second picture resolutions. The image processing device of claim 22, wherein the simulation unit provides a spatial data node in the at least one data node, corresponding to the spatial data node, the at least one connecting component comprising a space And a data connection component coupled between the spatial data node and the at least one diffusion node. 36